A prospective analysis of factors influencing outcome after fundoplication

A prospective analysis of factors influencing outcome after fundoplication

A Prospective Analysis of Factors Influencing Outcome After Fundoplication By Thomas R. Weber St Louis, Missouri l Fundoplication remains a common ope...

431KB Sizes 13 Downloads 82 Views

A Prospective Analysis of Factors Influencing Outcome After Fundoplication By Thomas R. Weber St Louis, Missouri l Fundoplication remains a common operation in the braindamaged pediatric patient, but recent reports suggest a poor outcome in these patients. The factors that might be associated with complications or recurrence after fundoplication have not been extensively examined. Fifty-six brain-damaged children, aged 6 months to 12 years, with documented gastroesophageal (GE) reflux underwent preoperative nutritional evaluations (percentage of ideal weight, albumin, nutrition risk index [NM]) and documentation of medications (dexamethasone for bronchopulmonary dysplasia) before standard Nissen fundoplication. Hospital stay, intensive care unit (ICU) stay, and time on ventilator, as well as major postoperative complications (wound infection/dehiscence, pneumonia) were prospectively analyzed. Survival and recurrence rates 1 to 3 years postoperatively were also assessed. Eighty-two percent of patients were cgO% ideal weight, and 50% had NRI < 90 (normal = 100) and 29% had albumin < 3.5 g/dL. Albumin < 3.5 was significantly (P c .Ol) associated with prolonged hospitalization (26.8 + 2.2 versus 15.1 + 1.1 days) and IClJ stay (13.8 + 1.0 versus 4.4 + .5 days) and time on ventilator (8.0 + 1.0 versus 1.8 + .4 days). NRI c 90 showed similar significant differences (P c -01). Ideal body weight < 90% was not significant. Major complications developed in 54% of patients; only two or more preoperative nutritional deficiencies, or a nutritional deficiency plus dexamethasone were significantly associated (P c .Ol). Recurrence occurred in 21% of patients and was significantly correlated with preoperative dexamethasone alone (P < .Ol), and especially when dexamethasone plus a nutritional deficit were present (low albumin, P < .OOl; low NRI, P < .005). No factor correlated with survival. These data show that preoperative nutritional status greatly affects short- and long-term results after fundoplication, especially when steroids are also administered, suggesting that preoperative nutritional support might benefit these patients. Copyright o 1995 by W. B. Saunders Company

INDEX

WORDS:

Gastroesophageal

reflux,

fundoplication.

F

UNDOPLICATION for gastroesophageal (GE) reflux refractory to medical therapy has become one of the more common operations performed in the pediatric age group. Many of these children have cerebral palsy or other devastating neurological disorders that have resulted in significant malnutrition, poor weight gain, and frequent pulmonary infections that may make them risky operative candidates. Several studies have recently called attention to the poor outcome after fundoplication in the braindamaged child,‘J but there have been few attempts to identify risk factors that may lead to postoperative complications or recurrence in this group of patients. The present study is an attempt to identify these

JournaloffedtatrtcSurgety,

Vol30,No

7(July),1995.

pp 1061-1064

factors prospectively in a group of brain-damaged infants and children undergoing fundoplication. MATERIALS

AND METHODS

From 1991 to 1993,56 infants and children with severe neurological impairment underwent fundoplication for gastroesophageal reflux. Forty children had cerebral palsy from birth asphyxia or intrauterine events, and 8 had suffered head injury from trauma. Six patients had hydrocephalus secondary to intraventricular hemorrhage, and 2 others had metabolic disorders that resulted in severe mental retardation. The patients were 6 months to 12 years of age. Each patient had gastroesophageal reflux documented by upper gastrointestinal contrast study and/or 12 to 24 hour continuous esophageal pH momtoring. A pH study was considered positive if the pH remained less than 4 for more than 10% of the study period, or if there were more than five episodes of pH less than 4 in a 1Zhour period. Each patient underwent a trial of medical therapy, usually including Ha blockers, prokinetic agents, and antacids for 2 to 8 weeks before referral for fundoplication. Forty-five of the 56 patients had suffered complications related to reflw before fnndoplication (recurrent aspiration pneumonia, esophageal stricture, esophageal hemorrhage), and the patients with the more serious or life-threatening complications tended to be referred earlier for fundoplication. Each patient underwent preoperative nutritional assessment within 1 week before Nissen fundoplication. This included measurement of weight, serum albumin, and calculation of a nutritional risk index (NRI), as previously described.3 This formula uses serum albumin and recent weight loss ratio for calculation. Although the NRI formula was originally described for adult patients, we have found it to be reliable in the pediatric patient as well (unpublished data). The ideal weight used in a ratio with preoperative weight was corrected for length because of the known tendency of brain-damaged children to have short length caused by limb atrophy and scoliosis.4 For each patient entered into the study, the following preoperative variables were recorded: percentage of ideal weight, serum albumin, NRI, and documentation of preoperative medications (dexamethasone). After performance of standard Nissen fundoplication, the following parameters were followed in the immediate postoperative period: Wound infection or dehiscence, urinary infection, pneumonia, prolonged atelectasis, time on ventilator postFrom the Division of Pediatric Surgery, St Louis University School of Medicine, and Cardinal Glennon Children’s Hospital, St Louis, MO. Presented at the 1994 Annual Meeting of the Section on Surgery of the American Academy of Pediatrics, Dallas, Texas, October 21-23, 1994. Address reprint requests to Thomas R. Weber, MD, Cardinal Glennon Children’s Hospital, 1465 S Grand Blvd, St Louis, MO 63104. Copyright o 1995 by W.B. Saunders Company 0022-3468/95/3007-0032$03.0010

1061

operatively, time in intensive care unit (ICU), and total hospital stay. In addition, survival and recurrence of gastroesophageal reflux was assessed for 1 to 3 years postoperatively. Recurrence was suspected when vomiting, substernal pain, or esophageal hemorrhage occurred in the postoperative period, and was confirmed by contrast study or pH study, using the same criteria as preoperatively. Correlation between preoperative risk factors and postoperative complications, recurrence rates, survival, and ventilator, and ICU and hospitalization durations were performed using a statistical software package and personal computer. Univariate analysis was performed by Pearson’s x2 with Yates’ correction or Fisher’s exact test. Stepwise logistic regression test was used for binomial data, complication versus no complication. RESULTS

All 56 patients survived the operative procedure and were eventually discharged. Forty-six of the 56 patients (82%) were < 90% ideal weight, and 28 of 56 (50%) had NRI score < 90 (normal = 100) and 16 of 56 (29%) had serum albumin < 3.5 g/dL. Eight of the 56 patients (14%) died within 3 years postoperatively of causes unrelated to the operation (seizure disorder, 3; hydrocephalus/brain atrophy, 2; pneumonia, 2; auto accident, 1). Both children who died with pneumonia had intact fundoplications at the time of death. No preoperative or postoperative factor correlated significantly with survival. Clinically apparent recurrent gastroesophageal reflux developed in 12 of 56 (21%) of patients, confirmed by upper gastrointestinal contrast study and/or pH monitoring. The development of recurrent reflux was significantly correlated with preoperative dexamethasone therapy (all patients on dexamethasone had recurrence, P < .Ol), dexamethasone plus albumin < 3.5 g/dL (90% had recurrence, P < .OOl), and dexamethasone plus NRI < 90 (90% had recurrence,

P < .005). The effects of preoperative serum albumin levels and NRI on the durations of hospital stay, ICU stay, and time on ventilator are shown in Figs 1 and 2. Both serum albumin and NRI reliably predicted the need Hospital Stay (PC 01)

DZZZJ 0

AlbumIn AlbumIn

< 3 5gidl > 3 5g/dl

Hospital Stay (PC 01)

ICU Stay (PC 01)

Fig 2. The effects of normal (NH) on length of hospital stay, fundoplication.

The effects of normal and abnormal albumin levels on length stay, ICU stay, and time on ventilator after fundoplication.

NRI < 90

0

NRI > 90

Time on Ventdator (PC 01)

and abnormal nutritional risk ICU stay, and time on ventilator

index after

DISCUSSION

The influence of preoperative risk factors on surgical outcome is a topic of considerable interest in both adult and pediatric patients.5-7 In an attempt to reduce the need for prolonged intensive care stays and hospitalizations, there have also been attempts to manipulate the preoperative status of the patient to reduce the intraoperative and postoperative risks, with variable success.3,5,8-1*Many of these studies have focused on improving the nutritional status of Table

1. Incidence

and

Significance

Fundoplication, Nutritional

Fig 1. of hospital

EZZZi

for prolonged ventilator, ICU, and hospitalization durations postoperatively. On the other hand, body weight < 90% of ideal weight was not predictive in this regard. Major complications (prolonged atelectasis, urinary infection, wound infection/dehiscence, pneumonia) developed in 30 of 56 (54%) of patients. No single nutritional measure or the use of dexamethasone alone was predictive of postoperative complications. However, albumin < 3.5 g/dL plus NRI < 90, or dexamethasone plus albumin < 3.5 g/dL or dexamethasone plus NRI < 90 together were associated with a greater incidence of postoperative complications (Table 1).

ICU Stay (P<.Oj)

L/

Ft. WEBER

30

After

Time on Ventilator (PC 01)

THOMAS

Vanable All values normal (A) AlbumIn i 3.5 g/dL (B) NRI < 90 (C) Preoperative A + B (above) A + C (above) B + C (above)

dexamethasone

of Postoperative in Relation

Complications

to Preoperative

Variables Incidence of Compllcatlons 1%)

P Value

4.1 6.2 5.5

NS NS

4 15.6 21 19.5

NS <.Ol <.Ol <.Ol

OUTCOME

AFTER

FUNDOPLICATION

1063

the malnourished adult surgical candidate before operation. However, there are few documented attempts to use similar nutritional assessments and interventions in the pediatric surgical patient. In the present study, a relatively homogeneous group of patients (brain-damaged children) with a high likelihood of malnourishment and who were all about to undergo the same operation (fundoplication) was selected to prospectively study the effects of preoperative nutrition or steroid administration on the short- and long-term results of the operative procedure. Although the group of patients is admittedly small, significant trends appeared as the data were analyzed that suggested a definite correlation between preoperative nutritional condition and eventual outcome. These data suggest that preoperative nutritional state as measured by serum albumin and NRI levels may predict the risk for postoperative complications and the risk of later recurrence of gastroesophageal

reflux. In addition, preoperative steroids, usually used in these patients for chronic pulmonary disorders such as bronchopulmonary dysplasia, are an additional risk factor, especially when combined with nutritional deficiency. The present study therefore suggests that preoperative aggressive nutritional intervention might reduce the incidence of complication in this group of patients, administered either enterally or parenterally. Enteral supplements are efficacious and would probably be cost effective,4 but obviously must be used cautiously in patients with a significant propensity for emesis and serious complications related to gastroesophageal reflux. The use of nasoduodenal or jejunal tubes might be safer in this setting. It would appear to be especially important to optimize nutritional status in children receiving steroids, because steroid medication and nutritional deficits seem to be additive with regard to increasing risk of complications and recurrence in this series.

REFERENCES 1. Smith CD, Otherson HB, Gogan NJ, et al: Nissen fundoplication in children with profound neurologic disability. High risks and unmet goals. Ann Surg 213:654-659, 1992 2. Tunnel1 W, Smith E, Carson J: Gastroesophageal reflux in childhood: The dilemma of surgical success.Ann Surg 197:560-565, 1983 3. The Veterans Affairs Total Parenteral Nutrition Cooperative Study Group: Perioperative total parenteral nutrition in surgical patients. N Engl J Med 325:525-532, 1991 4. Sanders KD, Cox K, Cannon R, et al: Growth response to enteral feeding by children with cerebral palsy. JPEN 14:23-26, 1990 5. Dempsey DT, Mullen JL, Buzby GP: The link between nutritional status and clinical outcome: can nutritional intervention modify it? Am J Clin Nutr 47:352-356.1988 6. Velanovich V: The value of routine preoperative laboratory

testing in predicting postoperative complications: A multivariate analysis. Surgery 109:236-243, 1991 7. Braga M, Baccari P, Scaccabarozzi S, et al: Prognostic role of preoperative nutritional and immunological assessment in the surgical patient. JPEN 12:138-142,1988 8. Smith RC, Hartemink R: Improvement of nutritional measures during preoperative parenteral nutrition in patient selected by the prognostic nutritional index: A randomized controlled trial. JPEN 12:587-591,1988 9. Belantone R, Doglietto GB, Bossola M, et al: Preoperative parenteral nutrition in the high risk surgical patient. JPEN 12:195197,1988 10. Detsy AS, Baker JP, O’Rourke K, et al: Perioperative parenteral nutrition: A meta-analysis. Ann Int Med 107:195-203, 1987 11. Megmid MM, Campos AC, Hammond WG: Nutritional support in surgical practice. Am J Surg 159:427-443, 1990

Discussion B. Hutis (Boston, MA): If you do the nutritional intervention before operation, what is the end point and how long does it take? T.R. Weber (response): I don’t know at this point. I would suspect that enteral supplementation through nasoduodenal tubes at home would be the most efficacious from a financial standpoint. J.L. Grosfeld (Indianapolis, IN): Did you include the use of gastrostomy tubes and early feeding in patients in the evaluation of your post operative parameters and your complication rates? In your patients who had adverse factors, is there some

correlation as to whether they had a higher rate of breakdown than the others? T.R. Weber (response): All of these patients had feeding gastrostomies placed. Their diets through the gastrostomy were begun when it was believed to be appropriate, in general 3 to 5 days after the operation. We did not find a correlation between recurrence and either early feedings or the preoperative evaluation. B.M. Rodgers (Charlottesville, VA): There must have been people in that group that had been nutritionally repleted before your Nissen. Did you separate out

1064

those individuals to see if preoperative support would change their outcome?

THOMAS

nutritional

T.R. Weber (response): No.

S.G. Jo&y (Las Vegas, NE): I think your paper showed an increase in wound infections and other complications after fundoplication in this very difficultto-treat group. I need to know a little bit more about how you have defined reflux not so much preoperatively but postoperatively, because you are claiming that nutritional factors are associated with an increased breakdown in the fundoplication. Did you do barium swallows or esophageal pH monitoring? What methods do you use? What was your range of normals? This is particularly important because most of the children that I see with recurrent symptoms of reflux do not have recurrent reflux, but rather gastric emptying abnormalities. Perhaps you are drawing the wrong conclusion. T.R. Weber (response): That’s a good point. We did not routinely look for reflux postoperatively. We relied on the clinical presentation of recurrent vomiting. Keep in mind these are all severely braindamaged children, so we were often unable to obtain symptoms of reflux. Clinically if they developed recurrent vomiting, then they were studied. There certainly

R. WEBER

could be a large population that have recurrent reflux that are not vomiting. But if the children developed vomiting or recurrent aspiration-type symptoms, then they were studied. S.G. Jo&y (Las Vegas, NE): Just to add to what I said, even most children with recurrent vomiting after fundoplication do not have recurrent reflux. T.R. Weber (response): All patients who had recurrent vomiting were then studied and found to have recurrent reflux. S.G. Jolley (Las Vegas, NE): By what methodology? T.R. Weber (response): By pH studies and/or barium swallows. J. Raffensperger (Chicago, IL): This is not a question, just a statement. At our hospital, after reviewing this whole situation with reflux in brain-damaged children, when we see a child lying in a bed in the fetal position who can not talk and can not eat, we do everything we can to discourage an operation. We do not improve their life-style, we do not improve their nutrition. All we do is succumb to the referring doctor’s desire to do something. And I think it is wrong. We should just say that the operation does not do a thing and not do it.